Adaptive Antioxidant Response of Manganese-SuperoxideDismutase Following Repetitive UVA Irradiation

Arndt Poswig, Jutta Wenk, Peter Brenneisen, Meinhard Wlaschek, Christina Hommel, Gudrun Quel,Katrin Faisst, Joachim Dissemond, Karlis Briviba,* Thomas Krieg, and Karin Scharffetter-KochanekDepartment of Dermatology, University of Cologne, Germany; *Department of Physiological Chemistry, I Heinrich Heine University of Dusseldorf,Germany

In response to the attack of reactive oxygen species,the skin has developed a complex antioxidant defensesystem including among others the manganese-superoxide dismutase (MnSOD). MnSOD dismutatesthe superoxide anion (O2·–) derived from the reductionof molecular oxygen to hydrogen peroxide (H2O2),which is detoxified by glutathione peroxidase to waterand molecular oxygen. We have addressed the questionwhether MnSOD is inducible upon UVA irradiationand whether repetitive UV exposure, as practiced forthe light-hardening during phototherapy of variousphotodermatoses, can even enhance the adaptive anti-oxidant response. Single exposure of four differentstrains of fibroblasts to UVA irradiation resulted in adose- and time-dependent increase in specific MnSODmRNA levels. Interestingly, repetitive UVA exposureat days 1, 2, and 3 at a dose rate of 200 kJ per m2

resulted in a 5-fold induction of specific MnSODmRNA levels following the third UVA exposure.

The skin is always in contact with oxygen and now isincreasingly exposed to ultraviolet (UV) irradiation.Therefore, the risk of photooxidative damage of theskin induced by reactive oxygen species (ROS) hasincreased substantially (Darr and Fridovich, 1994).

The term ROS collectively includes oxygen-centered radicals suchas the superoxide anion (O2·

–) and the hydroxyl radical (HO·), butalso some nonradical species, such as hydrogen peroxide (H2O2)and singlet oxygen (1O2), among others, all being produced in theskin upon UV irradiation. A complex antioxidant defense systemhas evolved in the skin and protects against ROS (Fridovich, 1989;Shindo et al, 1993); however, UV-generated ROS at least at highlevels may substantially compromise the antioxidant defense of theskin (Witt et al, 1993; Biesalski et al, 1996), thus tilting the balancetowards a prooxidant state (Sies, 1986). The resulting oxidativestress causes damage to cellular components and changes the patternof gene expression, finally leading to skin pathologies such as skincancer, phototoxicity, and photoaging (Oikarinen et al, 1985;

Manuscript received April 21, 1998; revised September 3, 1998; acceptedfor publication October 7, 1998.

Reprint requests to: Dr. Karin Scharffetter-Kochanek, Department ofDermatology, University of Cologne, Joseph-Stelzmann Str. 9, 50931Cologne, Germany.

Abbreviation: MnSOD, manganese-superoxide dismutase.

0022-202X/99/$10.50 · Copyright © 1999 by The Society for Investigative Dermatology, Inc.

13

Similar results were obtained for MnSOD activity.This adaptive response in terms of upregulation ofthe antioxidant enzyme MnSOD correlates with theprotection against high UV doses, if cells were pre-exposed to sublethal UV doses. Importantly, MnSODsubstantially differed between the tested individuals inboth mRNA and activity levels. Taken together, wehere provide evidence for the increasing induction ofMnSOD upon repetitive UVA irradiation that maycontribute to the effective adaptive UVA responseof the skin during light hardening in phototherapy.Interindividual differences in the inducibility ofMnSOD might account for differences in the suscepti-bility to develop photodermatologic disorders relatedto photosensitivity, photoaging, and skin cancer. Themolecular basis for interindividual differences in theinducibility of antioxidant enzymes remains to beelucidated. Key words: adaptation/MnSOD/skin/UVresponse. J Invest Dermatol 112:13–18, 1999

Gallagher et al, 1989; Urbach, 1989; Kligman, 1992; Scharffetter-Kochanek et al, 1997). As shown earlier, UV-generated ROSare able to induce the synthesis and activity of various matrix-metalloproteinases responsible for the breakdown of dermal inter-stitial collagen and other connective tissue components in vitro andin vivo (Scharffetter et al, 1991; Scharffetter-Kochanek et al, 1992;Wlaschek et al, 1994, 1995; Brenneisen et al, 1998).

Whereas high UV doses and ROS levels are known to overwhelmthe antioxidant enzymatic defense system, clinical studies haveshown that repetitive low UV doses prevent phototoxicity andUV-induced pathologies upon subsequent higher UV doses (Ruckeret al, 1991), indirectly suggesting that the skin – apart from othermajor protection mechanisms like inducible melanogenesis – ispossibly equipped with an inducible adaptive antioxidant defenseresponse.

The individual antioxidant enzymes are located in specificsubcellular sites and reveal substrate specificity. Among these,manganese-superoxide dismutase (MnSOD) has been the subjectof particular interest in terms of an inducible antioxidant defense,because it is located in the mitochondria and is in the first line ofdefense against superoxide radicals produced as a by-productof oxidative phosphorylation, upon UV irradiation and duringinflammatory processes. Interestingly, specific mRNA levels andMnSOD activity were found to be upregulated upon γ-irradiationin human lung fibroblasts (Akashi et al, 1995). The induction ofMnSOD can be mediated by its substrate, the superoxide anion

14 POSWIG ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

itself, and various cytokines such as IL-1 and TNF-α (Akashi et al,1995). The presence of MnSOD appears to be involved in processeslike tumor suppression and cellular differentiation (Harris et al,1991; Church et al, 1993; St. Clair et al, 1994; Akashi et al, 1995;Sato et al, 1995). Superoxide anions are dismutated by MnSOD tohydrogen peroxide, which is subsequently detoxified by catalaselocated in peroxisomes or by glutathione peroxidase located inmitochondria and the cytosol. Isolated deficiencies of superoxidedismutases as seen for the copper-zinc superoxide dismutase (Cu/ZnSOD) in patients suffering from amyotrophic lateral sclerosisand in mice completely deficient for MnSOD, are associated withpremature aging, neurodegeneration, and death (Li et al, 1995;Rosen et al, 1993; de Haan et al, 1996). Recently, we wereable to show that isolated overexpression of MnSOD by stabletransfection revealed specific resistance to the superoxide anion(O2·

–) generating agent paraquat (Wenk, personal communication).Here, we provide several lines of evidence for the protective roleof an inducible adaptive response of MnSOD upon single andrepetitive UVA irradiation. This adaptive response of an antioxidantenzyme like MnSOD was related to the protection against thecytotoxic effect of high UVA doses. Importantly, MnSOD substan-tially differed among eight tested individuals on both mRNA andactivity levels, which may account for differences in the susceptibil-ity to develop photodermatologic disorders related to phototoxicity,photoaging, and skin cancer.

MATERIALS AND METHODS

Reagents Glutathione, t-butyl-hydroperoxide, glutathione reductase,xanthin oxidase, and xanthin were obtained from Sigma (Deisenhofen,Germany) and Boehringer (Mannheim, Germany). The human MnSODprobe used was a 737 bp EcoRI/XbaI fragment of human MnSOD. A24-mer oligonucleotide (59-ACG GTA TCT GAT CGT CTT CGAACC-39) (Carlson et al, 1993) for the 18S rRNA was used (Pharmacia,Freiburg, Germany).

Cell cultures Human dermal fibroblasts were obtained from outgrowthfrom skin biopsies as earlier published (Fleischmajer et al, 1981). Cellswere maintained in Dulbecco’s modified Eagle’s medium (GIBCO BRL,Eggenstein, Germany) supplemented with 10% fetal calf serum (Biochrom,Berlin, Germany), glutamine (2 mM), penicillin (400 U per ml), andstreptomycin (50 mg per ml) at 37°C in a humidified atmosphere of 5%CO2 and 95% air.

Detection of MnSOD activity Fibroblast monolayers were washed,collected, and homogenized by sonification. MnSOD activity was measuredspectrophotometrically using the nitroblue tetrazolium reduction method(Fridovich, 1989). MnSOD activity was differentiated from Cu/ZnSODby its resistance to 5 mM cyanide.

Cytotoxicity assay Cytotoxicity was monitored at 24 h after irradiationat different UV doses. 3-(4,5-di-methylthiazol-2-yl)-2–5-diphenyltetra-zolium bromide was used for the quantitation of living metabolically activecells (Green et al, 1984). Mitochondrial dehydrogenases metabolize 3-(4,5-di-methylthiazol-2-yl)-2–5-diphenyltetrazolium bromide to a formazandye that is measured photometrically. Cytotoxicity was calculated aspercentage of formazan formation in cells subjected to different treatmentmodalities.

RNA extraction and northern blot analysis Total RNA was isolatedand analyzed by northern blots using specific cDNA probes or oligonucleo-tides for human MnSOD and 18S rRNA for sequential hybridization.Briefly, after extraction of total RNA, equal amounts (5–10 µg per lane)were fractionated by size on a 0.9% 2.2 M formaldehyde/agarose gel andblotted to nitrocellulose filters (Schleicher & Schuell, Dassel, Germany).Hybridizations were performed using denaturated 32P-labeled cDNAprobes. For 39-end labeling of the 24-mer 18S rRNA probe, 5 3 terminaldeoxynucleotidyl transferase buffer (0.5 M potassium buffercacodylate,pH 7.2, 10 mM CoCl2, 1 mM dithiothreitol), 10 pmol 39-ends (80 ngDNA of the 24-mer), 1.5 µCi α-32PdCTP per µl reaction volume, and1 U terminal deoxynucleotidyl transferase (GIBCO BRL) per ml wereincubated at 37°C for 1 h. Densitometric analysis was performed using theScanPackII system (Biometra, Gottingen, Germany).

Light source and UV irradiation The cells were irradiated at a distanceof 40 cm using a high intensity halogen metallide UVA source (UVASUN

3000 equipped with the UVASUN safety filters) emitting wavelengths in340–450 nm range (Mutzhas, Munich, Germany) (Lehmann et al, 1986).The spectral distribution of the UVASUN source was determined with aBeckman UV 5270 spectral photometer. The incident dose at the surfaceof the cells was 66 mW per s. Dose rates were monitored with acombined UVA/UVB ultravioletmeter (Centra-UV-dosimeter, Osram,Munich, Germany) (Mutzhas et al, 1981). During irradiation, cells wereincubated in phosphate-buffered saline and maintained at 37°C in athermostatically controlled water bath. Following irradiation, phosphate-buffered saline was replaced by fresh medium with 10% fetal calf serumand the cells were incubated for various periods of time. Applied UVAdoses were in the range between 50 and 300 kJ per m2. Even the highestdose of 300 kJ per m2 represents a dose easily acquired during 3 h sunexposure in June at latitude 40°N (Bruls et al, 1984; Kligman and Kligman,1986), thus showing the physiologic relevance of the applied doses.

RESULTS

UVA irradiation results in a dose- and time-dependentinduction of MnSOD on RNA and protein level In orderto study the effect of different UVA doses on the synthesis andactivity of MnSOD, we subjected total RNA isolated at differenttime points after UVA irradiation to northern blot analysis. In aparallel set of experiments, homogenized fibroblast monolayerswere prepared for spectrophotometric determination of MnSODactivity using the nitroblue tetrazolium reduction method.

We found a dose- and time-dependent increase in specificMnSOD mRNA levels. While already at an UVA dose of 50 kJper m2 both MnSOD mRNA species of 4 and 1 kb were slightlyincreased at 3 h post-irradiation, the most prominent increaseoccurred 1 h post-irradiation at an UVA dose of 300 kJ per m2,with a maximal 2-fold induction at 9–12 h and a subsequentdecline to almost basal levels 24 h post-irradiation. Interleukin-1β,a well-known MnSOD-inducing cytokine, served as a positiveinternal control in all experiments (Fig 1). We then determinedwhether the induction of MnSOD on RNA level is followed byan increase in the activity of MnSOD. There was no increase inMnSOD activity at any studied time point upon UVA irradiationat a dose of 50 kJ per m2; however, following UVA irradiation atdoses of 150 kJ per m2 or 300 kJ per m2, a biphasic dose-dependentincrease was observed. The first smaller peak occurred 3 h post-irradiation with an increase in MnSOD activity to 140% at a UVAdose of 150 kJ per m2 and to 120% at a UVA dose of 300 kJ perm2, compared with the mock treated control that was set as 100%.A second more impressive peak in MnSOD activity was observed12 h post-irradiation with an increase to 160% at a dose of 150 kJper m2 and to 170% at a dose of 300 kJ per m2 compared withthe mock treated control. At 24 h post-irradiation MnSOD activitylevels corresponded to constitutive basal levels (Fig 2).

Repetitive low-dose UVA irradiation further increasesspecific MnSOD mRNA and activity In a first attempt tostudy a potential adaptive antioxidant response in fibroblasts,fibroblast monolayer cultures were irradiated three times at a doseof 200 kJ per m2 with a 24 h incubation period between eachirradiation. MnSOD mRNA levels and activity were determinedafter the first, second, and third irradiation using northern blotanalysis and spectrophotometric determination of MnSOD activityfollowing standard procedures. After a single UVA dose of 200 kJper m2, there was an increase in specific MnSOD mRNA levelswith a maximal 1.5-fold induction. This induction was furtherincreased by 1.8- and 4.0-fold after the second and third UVAirradiation compared with values after the first UVA irradiation,suggesting that repetitive UVA irradiation results in a furtherenhancement of specific MnSOD mRNA levels (Fig 3). Thesefindings were also reflected on the level of MnSOD activity asshown in Fig 4. After the first UVA irradiation at a dose of 200 kJper m2, a 2-fold induction of MnSOD activity was observedcompared with the mock treated control, whereas fibroblast mono-layer cultures that had been subjected to two or three repetitiveUVA irradiations revealed a 2.3- and 3.3-fold increase in MnSODactivity compared with the mock treated control (Fig 4).

VOL. 112, NO. 1 JANUARY 1999 ADAPTIVE ANTIOXIDANT ENZYMATIC RESPONSE 15

Figure 1. UVA irradiation dose and time dependently induces thespecific MnSOD mRNA levels. Fibroblast monolayer cultures weresubjected to UVA irradiation at different doses (50, 150, 300 kJ per m2).Total RNA was isolated at 1, 3, 6, 9, 12, and 24 h post-irradiation andsubjected to northern blot analysis. Treatment of fibroblast monolayercultures with IL-1β (20 U per ml) served as positive control. Northernblots were sequentially hybridized with the cDNA probe for MnSOD andthe 24-mer oligonucleotide for 18S rRNA. The northern blot showsrepresentative data reproduced in two independent experiments.

Preirradiation of fibroblast monolayer cultures with lowUVA doses confers protection from the cytotoxic effect ofa subsequent high-dose UVA irradiation In order to studythe potential protective role of low-dose UV preirradiation priorto a cytotoxic high UVA dose, the viability of fibroblasts after threeexposures to low-dose UVA irradiation (200 kJ per m2) and a finalhigh-dose UVA irradiation (450 kJ per m2) (experimental groups1–3) was determined and compared with the viability of fibroblaststhat had been irradiated by a single high-dose UVA irradiation(experimental group 4) using the 3-(4,5-di-methylthiazol-2-yl)-2–5-diphenyltetrazolium bromide test. There was a significantprotection from the cytotoxic action of high-dose UVA irradiationin fibroblast monolayer cultures, which had been preirradiated onlyonce prior to exposure to the high UVA dose (Fig 5). The extentof protection from the cytotoxic effect of a single high-dose UVAirradiation could not be further increased by repetitive low-doseUVA irradiation, indirectly suggesting that the adaptive antioxidantresponse upon a single preirradiation at a low UVA dose of 200 kJper m2 is sufficient to prevent cytotoxicity by a subsequent highdose of UVA irradiation.

Evidence for interindividual differences in the constitutiveactivity and inducibility of MnSOD In order to study potential

Figure 2. UVA irradiation dose and time dependently inducesMnSOD activity. Fibroblast monolayer cultures were subjected todifferent doses of UVA irradiation (50, 150, 300 kJ per m2) and MnSODactivity was determined at different time points (0.5, 1, 3, 6, 9, 12, and24 h) post-irradiation using the nitroblue tetrazolium reduction method asdescribed in Materials and Methods. MnSOD activity was expressed aspercentage of the mock irradiated control. The experiments were performedin triplicate in three independent experiments with SD , 10%; *p ø 0.01compared with the MnSOD activity at 24 h post-irradiation.

Figure 3. Repetitive UVA irradiation results in a strongerenhancement of specific MnSOD mRNA levels. Fibroblast monolayercultures had been UVA irradiated at a dose of 200 kJ per m2 for a maximumof three times with a 24 h incubation period between each irradiation.Total RNA was isolated 24 h after the indicated irradiation (1, 2, and 3irradiation) and subjected to northern blot and densitometric analysis asdescribed in Materials and Methods. The northern blots were sequentiallyhybridized with the cDNA probe for MnSOD and the 24-meroligonucleotide for 18S rRNA. Densitometric data were standardized to18S ribosomal RNA and represent fold increase over control that was setat 1.0. A representative northern blot of two independent experimentsis shown.

interindividual differences in the constitutive synthesis andinducibility of MnSOD, fibroblast monolayer cultures were pre-pared by outgrowth from preputial biopsies from different healthyindividuals, and MnSOD activity was determined in mock treatedcontrols 12 h post-irradiation of fibroblasts at a dose of 200 kJ per

16 POSWIG ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure 4. Induction of MnSOD activity after repetitive UVAirradiation. Fibroblast monolayer cultures had repetitively been subjectedfor three times to UVA irradiation at a dose of 200 kJ per m2 with a 24 hincubation period between each irradiation. MnSOD activity wasdetermined as described in Materials and Methods 12 h after each irradiationand expressed as percentage of the mock treated control. The experimentswere performed in triplicate in three independent experiments withSD , 5%; *p , 0.005 compared with the mock treated controls thatwere set at 100%.

m2, as well as 12 and 24 h after exposure of fibroblast cultures torecombinant IL-1β at a concentration of 20 U per ml (Table I).There were strong interindividual differences in the constitutivesynthesis and the inducibility of MnSOD upon various stimuli.The spontaneous constitutive MnSOD activity among the studiedfibroblast cultures of eight different individuals differed 1.7-fold,with the lowest activity of MnSOD of 1.96 U per mg and thehighest activity of 3.36 U per mg. There are, however, mostdramatic interindividual differences following UVA irradiation witha maximal 3.8-fold difference in MnSOD activity in differentfibroblast strains. Similarly, IL-1β preincubation of individualfibroblast strains for 12 and 24 h resulted in differences in MnSODactivities ranging from 2.5 at 12 h post-treatment to 2.4 at24 h post-treatment. Strong interindividual differences in thespontaneous activity and inducibility of MnSOD may, thus, conferdifferences in the interindividual susceptibilty for the developmentof skin aging and cancer.

DISCUSSION

The incidence of nonmelanoma and melanoma skin cancers, byfar the most common forms of human cancer, has increasedsubstantially over recent decades, with annually new cases fornonmelanoma skin cancer in the U.S.A. between 600,000 (Kwaet al, 1992) and 1,200,000 (Miller and Weinstock, 1994). Thistendency has been noted worldwide (for review, see Scharffetter-Kochanek and Mauch, 1994). Also, the incidence of photoaginghas increased tremendously. The increasing incidence of skinmalignancies and photoaging is considered to be at least in partdue to UV-generated ROS. This is particularly relevant in termsof UV irradiation, which is predicted to increase on the earth’ssurface upon ozone depletion (Coldiron, 1992; Slaper et al, 1996).Therefore, an understanding of the antioxidant enzymes withinresident skin cells and the regulation upon exposure to UV-generated ROS is important for our understanding of UV-protective

Figure 5. Low-dose UVA preirradiation confers protection fromcytotoxicity of a subsequent high dose of UVA irradiation. Fourexperimental groups were defined, as indicated on the y axis. Inexperimental group 1, fibroblasts had been exposed once, in experimentalgroup 2, fibroblasts had been exposed twice, and in experimental group3, fibroblasts had been exposed three times to low doses of UVA irradiation(200 kJ per m2) prior to being exposed to a subsequent high UVA dose(450 kJ per m2). In a parallel experiment fibroblasts of experimental group4 were directly exposed to a single high UVA dose (450 kJ per m2).Viability of low-dose preirradiated and subsequently high-dose irradiatedfibroblasts (experimental groups 1, 2, and 3) and the viability of fibroblaststhat had been exposed to a single high dose (experimental group 4), wasdetermined using the 3-(4,5-di-methylthiazol-2-yl)-2–5-diphenyltetra-zolium bromide assay as described in Materials and Methods. Viability wasexpressed as percentage of the viability of mock treated fibroblasts that wasset 100%. The viability of fibroblasts of experimental groups 1, 2, and 3was compared with that of fibroblasts of experimental group 4. Theexperiments were performed in triplicate in three independent experimentswith SD , 8%; *p , 0.001.

Table I. Interindividual differences in the constitutive andinducible activity of MnSOD

MnSOD activity (U per mg)a

UVA IL-1β (20 U per ml)Age Mock (200 kJ per m2)

Donor (y) treated 12 h 12 h 24 h

KU 52 1.96 3.14 2.74 4.31ER 5 3.22 3.7 6.44 8.66PL 9 3.28 3.64 6.78 7.64GO 59 3.6 5.43 n.d. n.d.MB 5 2.6 1.95 3.5 3.9ED 10 2.8 2.0 3.0 3.5FK 6 3.04 4.0 n.d. n.d.FD 9 3.36 7.52 n.d. n.d.

aMnSOD activity was determined in fibroblast monolayer cultures from differenthealthy individuals after treatment with UVA or IL-1β at the indicated dose andconcentration. MnSOD activity is expressed in U per mg. n.d., not done. Theexperiments were performed in three independent experiments with SD , 10%.

mechanisms in skin, urgently required for proper individual riskassessment and the protection of sun-exposed skin.

The overall aim of this study was therefore to define adaptiveantioxidant enzymatic mechanisms that confer protection to residentcells against the oxidative attack following UVA irradiation.

We found that UVA irradiation of fibroblast monolayer culturesin vitro resulted in a significant increase in MnSOD on both mRNAand protein levels. The administered UVA doses in our experiments

VOL. 112, NO. 1 JANUARY 1999 ADAPTIVE ANTIOXIDANT ENZYMATIC RESPONSE 17

are physiologically relevant and can be easily acquired (Bruls et al,1984; Kligman and Kligman, 1986), thus underlining the relevanceof our findings. Furthermore, we found that repetitive low-doseUVA irradiation could even dramatically enhance the synthesis andactivity of MnSOD, and that this induction is related to a substantialprotection against the cytotoxic effect of a subsequent high-doseUVA insult.

Earlier, it was reported that chronic UV exposure of hairlessmice for 2 h per day with a source mainly emitting UVA including2% UVB, resulted in a significant increase in SOD activity that,however, following continued irradiation for 24 wk, substantiallydecreased below the level of mock treated animals (Maeda et al,1991), whereas glutathione peroxidase-activity remained raised.These results suggest that chronic UV exposure for months, evenat suberythermal doses, may compromise the SOD-dependentantioxidant defense. It remains to be determined whether short-term repetitive UVA irradiation – as observed in our in vitroexperiments – may also result in an increase in MnSOD activityin human skin in vivo and whether repetitive irradiation overmonths – similar to the findings in mice – will compromise SODactivity in human skin. The MnSOD is the first line of defense inthe protection against the attack of the superoxide anion and revealstumor suppressing properties (Church et al, 1993). Given that thegene of the MnSOD is deleted in some but not all melanomas,our finding of an adaptive and superinducible response of theMnSOD upon repetitive UVA irradiation is particularly interestingand may have in vivo relevance for the protection of cellularstructures and functions. In fact, loss of cellular redox homeostasis,due to a failure of the antioxidant enzymes, as occurs in severesunburns, represents a major risk factor in the etiology of melanoma(MacKie and Aitchison, 1982; Setlow et al, 1993; Autier et al,1994; Klein-Szanto et al, 1994; Holly et al, 1995; Donawho andWolf, 1996).

Similar to nonmelanoma skin cancer, melanoma also undergoesa multistage development towards the fully malignant phenotype.A failure of proper detoxification of ROS by antioxidant enzymesresults in an increase of the load of ROS that had been shown tobe involved in all three stages of carcinogenesis (Cerutti et al,1994), and is produced by both the UVA and the UVB componentsof sunlight (Masini et al, 1994). As well as causing permanentgenetic changes involving protooncogenes and tumor suppressorgenes, ROS activate cytoplasmic signal transduction pathwaysthat are related to growth, differentiation, senescence, and tissuedegradation (Cerutti, 1994; Brenneisen et al, 1998). The adaptiveinduction of the MnSOD upon UVA irradiation definitely con-tributes to the detoxification of the superoxide anion (O2), whichif not detoxified could at least partly drive the Fenton reactionresulting in the generation of the highly aggressive hydroxyl radical(HO·) (Darr and Fridovich, 1994). The overall relevance ofMnSOD becomes particularly clear in MnSOD deficient micerecently generated by gene targeting and homologous recombinanttechniques. Complete absence of MnSOD results in lethality ofnewborn mice due to defects in the oxidative phosphorylationpathway resulting in a dilatative cardiomyopathy as early as 10 dafter birth (Li et al, 1995).

Single UVA irradiation and to a significantly greater extentrepetitive irradiation for three times, resulted in a dramatic increasein the two specific 1.0 and 4.0 kb sized MnSOD mRNA speciesas well as in MnSOD activity. Whereas other investigators foundthat the half times of the MnSOD-specific transcripts substantiallydiffered (Melendez and Baglioni, 1993), we did not detect anydifference in the steady-state mRNA levels of both species uponUVA irradiation. It is still unclear whether both transcripts areequally required for the overall synthesis and activity of MnSOD(Melendez and Baglioni, 1993). MnSOD activity occurred in twopeaks upon UVA irradiation. Interestingly, this biphasic time coursein the induction of MnSOD activity has earlier been described byOberley et al (1987) following x-irradiation of mouse heart. Basedon the lack of 3[H]arginine incorporation, the nature of the firstpeak appears to be due to a MnSOD precursor protein that becomes

active upon still unidentified stimuli, whereas the second peak inMnSOD activity – as shown by the dramatic incorporation of3[H]arginine – in fact, requires new protein synthesis (Oberley et al,1987). The biphasic response in MnSOD activity may be ofphysiologic relevance in that the first peak may confer protectionagainst O2·

– early after the oxidative insult, whereas the secondpeak may protect cells from repetitive UVA irradiation or from asustained ROS-generating inflammation following UVA irradiationor other ROS-generating stimuli.

As to the regulation of the induced synthesis and activity ofMnSOD upon UVA irradiation, it is most likely that O2·

– and IL-1α and IL-1β, both released following UVA irradiation, qualify tomediate UVA effects (Kupper et al, 1987; Masini et al, 1994;Wlaschek et al, 1994). At least, both O2·

– and IL-1α and IL-1βare well-known stimuli of the MnSOD (Marklund, 1992; Akashiet al, 1995). Preliminary data, in fact, show that neutralizingantibodies against IL-1α and IL-1β can at least partly suppress theinduction of MnSOD activity following UVA irradiation (data notshown). It is, however, unresolved whether O2·

– and IL-1 representcompletely independent mechanisms or act as causally relatedsequential events. So far, there is some evidence that IL-1 is ableto induce the generation of ROS in fibroblasts (Meier et al, 1989;Lee et al, 1996).

Low-dose preirradiation definitely conferred protection from thecytotoxic effects of a subsequent high dose of UVA irradiation.Even though we only provide correlative evidence, it is most likelythat the induction of MnSOD at least in part contributes to thisprotection.

Our finding that strong interindividual differences exist interms of spontaneous activity and inducibility of MnSOD uponstimulation with UVA irradiation or IL-1β, may represent amolecular determinant conferring differences in the observed inter-individual susceptibility for the development of premature agingand skin cancer. Further elucidation of the adaptive superinducibleantioxidant defense systems in physiologic and pathologic conditionswill not only allow individual risk assessment but will also stimulatethe development of novel therapeutic strategies.

We thank Irene Smith (Genentech, San Francisco, CA) for supplying us with thecDNA clone for human MnSOD. Karin Scharffetter-Kochanek was a recipicientof a Heisenberg grant from the German Research Foundation (DFG).

REFERENCESAkashi M, Hachiya M, Paquette RL, Osawa Y, Shimizu S, Suzuki G: Irradiation

increases Manganese Superoxide Dismutase mRNA levels in human fibroblasts.J Biol Chem 270:15864–15869, 1995

Autier P, Dore JF, Lejeune F, et al: Recreational exposure to sunlight and lack ofinformation as risk factors for cutaneous malignant melanoma: results of anEuropean Organization for Research and Treatment of cancer (EORTC) case-control study in Belgium, France and Germany. Melanoma Res 4:79–85, 1994

Biesalski HK, Hemmes C, Hopfenmuller W, Schmid C, Gollnick HP: Effects ofcontrolled exposure of sunlight on plasma and skin levels of β-carotene. FreeRadic Res 24:215–224, 1996

Brenneisen P, Wenk J, Klotz LO, et al: Central role of ferrous/ferric iron in the UV-B irradiation-mediated signaling pathway leading to increased interstitialcollagenase [matrix-degrading metalloprotease (MMP)-1] and stromelysin-1(MMP-3) mRNA levels in cultured human dermal fibroblasts. J Biol Chem273:5279–5287, 1998

Bruls WAG, Slaper H, van der Leun JC, Berrens L: Transmission of human epidermisand stratum corneum as a function of thickness in the ultraviolet and visiblewavelengths. Photochem Photobiol 40:485–494, 1984

Carlson SG, Fawcett TW, Bartlett JD, Bernier M, Holbrook NJ: Regulation of theC/EBP-related gene gadd 153 by glucose deprivation. Mol Cell Biol 13:4736–4744, 1993

Cerutti P, Gosh R, Oya Y, Amstad P: The role of the cellular antioxidant defensein oxidant carcinogenesis. Environ Health Perspect 102:123–129, 1994

Church SL, Grant JW, Ridnour LA, Oberley LW, Swanson PE, Meltzer PS, TrentJM: Increased manganese superoxide dismutase expression suppresses themalignant phenotype of human melanoma cells. Proc Natl Acad Sci USA90:3113–3117, 1993

Coldiron BM: Thinning of the ozone layer: facts and consequences. J Am AcadDermatol 27:653–662, 1992

Darr D, Fridovich I: Free radicals in cutaneous biology. J Invest Dermtol 102:671–675, 1994

18 POSWIG ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Donawho C, Wolf P: Sunburn, sunscreen, and melanoma. Curr Opin Oncol 8:159–166, 1996

Fleischmajer R, Perlish JS, Krieg T, Timpl R: Variability in collagen and fibronectinsynthesis by scleroderma fibroblasts in primary culture. J Invest Dermatol 76:400–403, 1981

Fridovich I: Superoxide dismutase: an adaption to a paramagnetic gas. J Biol Chem264:7761–7764, 1989

Gallagher RP, Elwood JM, Yang CP: Is chronic sunlight exposure important inaccounting for increase in melanoma incidence? Int J Cancer 44:813–815, 1989

Green LM, Reade JL, Ware CF: Rapid colorimetric assay for cell viability: applicationto the quantitation of cytotoxic and growth inhibitory lymphokines. J ImmunolMeth 70:257–268, 1984

de Haan JB, Christiano F, Ianello RC, Bladier C, Kelner MJ, Kola I: Elevation inthe ratio of Cu/Zn-superoxide dismutase to glutathione peroxidase activityinduces features of cellular senescence and this effect is mediated by hydrogenperoxide. Hum Mol Genet 5:283–292, 1996

Harris CA, Derbin KS, Hunte-McDonough B, Krauss MR, Chen KT, Smith DM,Epstein LB: Manganese superoxide dismutase is induced by IFN-gamma andtumor necrosis factor or IL-1. J Immunol 147:149–154, 1991

Holly EA, Aston DA, Cres RD, Ahn DK, Kristiansen JJ: Cutaneous melanoma inwomen: I. Exposure to sunlight, ability to tan and other risk factors related toultraviolet light. Am J Epidemiol 141:923–933, 1995

Klein-Szanto AJP, Silvers WK, Mintz B: Ultraviolet radiation-induced malignantskin melanoma in melanoma-susceptible transgenic mice. Cancer Res 54:4569–4572, 1994

Kligman LH: UVA induced biochemical changes in hairless mouse skin collagen: Acontrast to UVB effects. In: Urbach F (ed.). Biological Responses to UltravioletRadiation. Overland Park: Valdemar Publishing, 1992, pp. 209–216

Kligman LH, Kligman AM: The nature of photoaging: its presentation and repair.Photodermatology 3:215–227, 1986

Kupper TS, Chua AO, Flood P, McGuire J, Gubler U: Interleukin 1 gene expressionin cultured human keratinocytes is augmented by ultraviolet radiation. J ClinInvest 80:430–436, 1987

Kwa RE, Campana K, Moy RL: Biology of cutaneous squamous cell carcinoma.J Am Acad Dermatol 26:1–26, 1992

Lee SF, Huang YT, Wu WS, Lin JK: Induction of c-Jun protooncogene expressionby hydrogen peroxide through hydroxyl radical generation and p60src tyrosinekinase activation. Free Rad Biol Med 21:437–448, 1996

Lehmann P, Holzle E, Kries van R, Plewig G: Lichtdiagnostische Verfahren beiPatienten mit Verdacht auf Photodermatosen. Zentralbl Haut 152:667–682, 1986

Li Y, Huang TT, Carlson EJ, et al: Dilated cardiomyopathy and neonatal lethality inmutant mice lacking manganese superoxide dismutase. Nature Genet 11:376–381, 1995

MacKie RM, Aitchison T: Severe sunburn and subsequent risk of primary cutaneousmalignant melanoma in Scotland. Br J Cancer 46:955–960, 1982

Maeda K, Naganuma M, Fukuda M: Effects of chronic exposure ultraviolet-Aincluding 2% ultraviolet-B on free radical reduction systems in hairless mice.Photochem Photobiol 54:737–740, 1991

Marklund SL: Regulation by cytokines of extracellular superoxide dismutase andother superoxide dismutase isoenzymes in fibroblasts. J Biol Chem 267:6696–6701, 1992

Masini V, Noel-Hudson MS, Wepierre J: Free-radical damage by UV orhypoxanthine-xanthine oxidase in culturred human skin fibroblasts: protectiveeffect of two human plasma fractions. Toxicol In Vitro 8:235, 1994

Meier B, Radeke HH, Selle S, Younes M, Sies H, Resch K, Habermehl GG:Human fibroblasts release reactive oxygen species in response to interleukin-1or tumor necrosis factor-α. Biochem J 263:539–545, 1989

Melendez JA, Baglioni C: Differential induction and decay of manganese superoxidedismutase mRNAs. Free Radic Biol Med 14:601–608, 1993

Miller DL, Weinstock MA: Nonmelanoma skin cancer in the United States:incidence. J Am Acad Dermatol 30:774–778, 1994

Mutzhas MF, Holzle E, Hofmann C, Plewig G: A new apparatus with high radiationenergy between 320 and 460nm: physical description and dermatologicalapplications. J Invest Dermatol 76:42–47, 1981

Oberley LW, St. Clair DK, Autor AP, Oberley TD: Increase in manganese superoxidedismutase activity in the mouse heart after X-irradiation. Arch Biochem Biophys254:69–80, 1987

Oikarinen A, Karvonen J, Uitto J, Hannuksela M: Connective tissue alterations inskin exposed to natural and therapeutic UV-radiation. Photodermatol 2:15–26, 1985

Rosen DR, Siddique T, Patterson D, et al: Mutations in Cu/Zn superoxide dismutasegene are associated with familial amyotrophic lateral sclerosis. Nature 362:59–62, 1993

Rucker BU, Haberle M, Koch HU, Bocionek P, Schriever KH, Hornstein OP:Ultraviolet light hardening in polymorphus light eruption – a controlled studycomparing different emission spectra. Photodermatol Photoimmunol Photomed8:73–78, 1991

Sato M, Sazaki M, Hojo H: Antioxidative roles of metallothionin and manganesesuperoxide dismutase induced by tumor necrosis factor alpha and interleukin-6. Arch Biochem Biophys 316:738–744, 1995

Scharffetter K, Wlaschek M, Hogg A, et al: UVA-irradiation induces collagenase inhuman dermal fibroblasts in vitro and in vivo. Arch Dermatol Res 283:506–511, 1991

Scharffetter-Kochanek K, Mauch C: Hauttumor-Bindegewebs-Interaktionen. In:Macher E, Kolde G, Bocker EB (eds). Jahrbuch der Dermatologie, Zulpich,Germany: Biermann-Verlag, 1994

Scharffetter-Kochanek K, Wlaschek M, Bolsen K, et al: Mechanisms of cutaneousphotoaging. In: Plewig G, Marks R (eds). The Enviromental Threat of the Skin,London: Martin Dunitz Publishers, 1992, pp. 72–82

Scharffetter-Kochanek K, Wlaschek M, Brenneisen P, Schauen M, Blaudschun R,Wenk J: UV-induced reactive oxygen species in photocarcinogenesis andphotoaging. Biol Chem 378:1247–1257, 1997

Setlow RB, Grist E, Thompson K, Woodhead AD: Wavelength effective in inductionof malignant melanoma. Proc Natl Acad Sci 90:6666–6670, 1993

Shindo Y, Witt E, Packer L: Antioxidant defense mechanisms in murine epidermisand dermis and their responses to ultraviolet light. J Invest Dermatol 100:260–265, 1993

Sies H: Biochemistry of oxidative stress. Angew Chemie 25:1058–1071, 1986Slaper H, Velders GJM, Daniel JS, de Griujl FR, van der Leun JC: Estimates of

ozone depletion and skin cancer incidence to examine the Vienna conventionachievements. Nature 384:256–258, 1996

St. Clair DK, Oberley TD, Muse KE, St Clair WH: Expression of manganesesuperoxide dismutase promotes cellular differentiation. Free Rad Biol Med16:275–282, 1994

Urbach F: Potential effects of altered solar ultraviolet radiation on human skin cancer.Photochem Photobiol 50:507–514, 1989

Witt EH, Motchnik P, Packer L. Evidence for UV light as an oxidative stressor inskin. In: Fuchs J, Packer L (eds). Oxidative Stress in Dermatology, New York:Dekker, 1993

Wlaschek M, Heinen G, Poswig A, Schwarz A, Krieg T, Scharffetter-KochanekK: UVA induced stimulation of fibroblast-derived collagenase/MMP-1 byinterrelated loops of interleukin-1 and interleukin-6. Photochem Photobiol59:550–556, 1994

Wlaschek M, Briviba K, Stricklin GP, Sies H, Scharffetter-Kochanek K: Singlet oxygenmay mediate the ultraviolet A induced synthesis of interstitial collagenase. J InvestDermatol 104:194–198, 1995